The aim of this randomized clinical trial is to measure the effectiveness of the Personalized Nursing LIGHT behavioral intervention model in decreasing addiction severity, reducing drug-using and sexual practices that are high-risk for HIV transmission, and improving perceived sense of well-being of active IV drug users (IVDUs) who are not currently in treatment and who self-present for medical care in community hospital emergency rooms. Previous exploratory/demonstration projects have indicated that the LIGHT Model is a promising intervention, both in reaching and serving hard-to-reach active IVDUs who are not in treatment and in reducing their high-risk AIDS behaviors. A sample of 4,320 IVDUs, who self-present to one of three participating hospital emergency rooms in Detroit, MI; Baltimore, MD; and Brooklyn, NY, and who provide informed consent, will be stratified by race, gender, and site and randomized to receive either standard care standard care enhanced by the Personalized Nursing LIGHT Model. All subjects will be interviewed at admission to the study, and at six and twelve months following treatment completion. The standard care group will to an informational tape about AIDS and AIDS prevention and will be given a prevention and referral packet. In addition, the enhanced care group will receive three additional sessions which will provide indepth personalized teaching and counseling using the LIGHT Model. Data will be analyzed using a four-factor MANOVA factorial design with repeated measures. The independent variable is type of care, i.e., LIGHT vs standard (two levels). The three moderator variables are race, gender, and site (3 x 2 x 3). MANOVA will also provide a test of differences between the two types of care over the three data collection points (initial,6 months, and one year) on the multiple dependent variables (addiction severity, high-risk drug use and sexual practices, and well being).